FLOOD09784 - Laserfiche WebLink

Home Browse Search

EM 1I1o--:t-t{08 1 Mareh l!l6O \- ,~ ler but the progressive average-lag method is simpler to apply and is particularly adaptable for pre- liminary estimates in actual operations or for relath'e evaluation in the case of e. system of proposed reseryolI'S. The progressive average-lag method was applied to the holdouts of eighteen potential reservoirs in the Kansas River basin in studying the July 1951 flood. In this case, the eighteen sites had been selected from a long list of potential reservoir sites. The routing study then was made to determine the relative effectiveness of each reservoir in reducing flood discharges at or near the principal damage centers in the basin, that is at Enterprise, Junction City, Wamego, Topeka, and Kansas City. It Was necessary first to develop natural discharge hydrographs at some of the stations. This was done with the aid of slope- area measurements of peak discharges and routings by' the coefficient method from main stem stations where hydrographs were relatively reliable and from tributary stations where in some cases hydrographs were developed from rainfall and unit hydrographs. Next some trial runs were made with the progressive average-lag method to determine the time periods of inflow to be averaged to obtain outflows from a reach. This period was found to be about three-fourths of the time of travel or lag period. The routing IDethod was then applied to hypothetical holdouts at each; cservoir assuming complete control during the period 10-15 July 1951. A table of routings was prepared for each reservoir according to the example of plate No. 12. Then discharge reductions for the period 12-15 July 1951 were entered on tables for each principal station, that for Kansas City being shown on plate No. 13. Computations of the type shown on plate No, 13 yielded hydrographs at the damage center which would result from the hypothetical operation of some or all of the reservoirs. The table indicates the amount each reservoir reduces tbe natural crest discharge. The table discloses also that the peak routed holdouts of some reservoirs occur after the time of natural flood peak and in other cases before the natural peak. Neither the holdout at the time of the flood crest nor the peak holdout at a time otber than the flood crest as an adequate measure of the effect of a reservoir at the damage center. Instead in the 1951 study the flow reduction averaged over the 72-hour period of highest discharge is defined as an index of reservoir effec- tiveness. For Kansas City, the 72-hour period covered the calendar daya 13-15 July 1951, at Topeka the period started at noon 12 July 1951, and so on. Reservoir effectiveness indices for these and other reservoirs are shown in plate No. 14 with reference to each of the damage centers. This type of table is very effecti'l'e in presenting the results of flood routings for reservoir planning studies. Space is provided in the table of plates Nos. 12 to 14 for the results of a critical storm transposition over the basin. In the case of the transposition, the natural hydrographs at the stations in the basin may first be determined by the combined use of rainfall-unit hydrograph analysis and flood routing, after which the holdouts may be routed in the manner shown in plate No. 12. II} the case of the Kansas River studies, routings of holdouts were made subsequently by the coeffi- cient method. Peak discharges of the hydrographs differed by an average of 7.5 per cent. The results of the two methods then were considered in adopting a. final discharge fot use in project reports. Although two independent determinations l're, usually better than one, one determina.tion will suffice in-many cases. FOR THE CRIEF OF ENGINEERS: W. P. LEBER, Cok>-Ml, Corps of Engineers, Eucutive. ;. 17 i